Spatial-temporal Analysis Of Crop Growth Parameters With Assimilating Remote Sensing Into Crop Growth Model

The assimilation of remote sensing observation into the numerical dynamic modelis one efficient way to obtain the surface parameter accurately and analyze the surfaceconditions on spatial-time scale. In this study, take crop growth model which is one ofnumerical dynamic model for example, I discuss the assimilation method andtechnology of remote sensing information and crop growth model, and analyze thecrop growth conditions on spatial-time scale based on this method.In this study, I select the cropland of ChangChun area of the Northeast Plain asthe experimental field, and collect the essential data such as hyper-spectral ASD dataabout several critical rice growing period, meteorological data, Biochemical data, andget MODIS data and quasi-synchronous CCD data. Based on these data, I extract theoptimal observed quantities of the assimilation of remote sensing information intocrop growth model, discuss the problem about the time selection of the assimilation,do the research of the method of the regionalization of the model parameter, build theregional assimilation frame of remote sensing information into crop growth model, dothe dynamic analysis at spatial-time scale on the rice growth situation of ChangChunarea based on the assimilation frame. In this paper, research work and conclusions areas follows:(1) Based on the PROSPECT+SAIL radiation transfer model analyzing thesensitivity of the vegetation index, selecting the vegetation index which is the mostsensitive to the coupling parameter LAI of the coupling modelWOFOST+PROSPECT+SAIL as the observed quantity of assimilating remotesensing information into crop growth model. The results indicate that the observedquantities MCARI1is better to reflect the variant of LAI, and as the observedquantities of the assimilation, it is more efficient than that of all band reflectance.(2) Scheduling the assimilation of different step length observed quantities,comparing the accuracy and the efficiency of the assimilation at different time scale,selecting the proper time scale of the assimilation. The results indicate that selectingthe time scale of the step length between10days and20days about the assimilation of the remote sensing information and WOFOST model is more appropriate.(3) Adopt the sensitivity analysis approach to select and improve currentvegetation index, obtaining spatial distribution values depended on the spatialheterogeneity characteristics parameter of the inversion model based on CCD andMODIS data. The results indicated that the GNDVI and NDWI2could do a betterinversion of chlorophyll and the moisture contents, all of R2are higher than80%.(4) After finishing the above research, based on PSO algorithm, building thepotential productivity regional assimilation frame of remote sensing information intothe crop growth model. The spatial-temporal analyses were done on the rice growthsituation of ChangChun area with the assimilation frame. The results indicate that theregional assimilation frame RS-WOPROSAIL realizes the continuous simulation ofcrop growth parameters at spatial-time scale. The analog LAI, WSO, and TAGP areproperly reflects the change of the rice growth situation at spatial-time scale.Main innovation points of this thesis are as follow: select the assimilationobserved quantity with radiation transfer model, and discuss the selection of theassimilation time scale of remote sensing observation; provide an efficient method tosolve the problem about increasing the scale of the model with the spatialheterogeneity characteristics parameters of the Multi-source remote sensing datainversion coupled model such as CCD and MODIS; based on PSO algorithm, buildthe regional assimilation frame of remote sensing, this method has widespreadapplicability, is very good at the mechanism and can be widely adopted on thespatial-time data model and analysis.The regional assimilation frame of remote sensing information into crop growthmodel based on PSO algorithm solves the problem about the temporal discontinuationof remote sensing and spatial discontinuation of the crop growth model, generates thedata set of the crop growth conditions, analyzes the crop growth condition onspatial-time scale with the dataset, and also provides some reference and ideas aboutthe modeling and analyzing depended on the spatial-time data.